As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
A barometer is an instrument that measures atmospheric pressure, the force per unit area exerted by the weight of the air above the sensor. The three common mechanisms are liquid column (mercury), elastic deformation (aneroid capsule or Bourdon tube), and solid-state piezoresistive or capacitive silicon. Handheld weather meters such as The BTMETER BT-100WM and The H HOLDPEAK 866B-WM combine a silicon pressure sensor with wind and humidity functions, while fixed stations like The Netatmo Smart Weather and The ECOWITT GW2001 report pressure alongside other weather variables. None of these instruments is a calibrated pressure standard, and none replaces a barometric altimeter certified for aviation or a leak detector for gas piping.
The weight of air, and why a column of mercury was the first answer
At sea level, the atmosphere presses down with a force of roughly 101,325 pascals, which is 1013.25 millibars or 29.92 inches of mercury. Those three numbers describe the same pressure in different units. A barometer is simply a device that compares the unknown atmospheric pressure against a known reference.
In a mercury barometer, the reference is a glass tube sealed at one end, filled with mercury, and inverted into a reservoir of mercury open to the air. The atmosphere pushes down on the reservoir surface, and the mercury column rises inside the tube until the weight of the column balances that push. At standard sea-level pressure the column stands about 760 millimetres tall. The instrument reads the height of the column, and that height is the pressure.
Mercury barometers are accurate and simple, but they are also fragile, toxic if broken, and sensitive to temperature because mercury expands when warm. A modern aneroid barometer replaces the liquid column with a sealed, partially evacuated metal capsule. As atmospheric pressure rises, the capsule compresses; as pressure falls, it expands. A linkage magnifies that tiny movement and drives a needle across a dial. The mechanism is mechanical, so it drifts and needs occasional adjustment against a known reference.
The third family is solid state. A silicon diaphragm is etched into a chip, and piezoresistive elements or a capacitive gap measure how far the diaphragm deflects under pressure. These sensors are small, cheap, and can resolve a fraction of a hectopascal, which is why they appear in handheld weather meters and phone-based stations. The trade-off is that absolute accuracy depends on factory calibration and on how well the sensor is isolated from temperature and mechanical stress.
What the number on the display actually represents
A barometer reports absolute pressure, the total pressure at the sensor port. That is different from gauge pressure, which is measured relative to the local atmosphere. A tyre gauge reads gauge pressure; a barometer reads absolute pressure. The distinction matters when you compare readings from different instruments, because a gauge pressure of zero corresponds to an absolute pressure near 1013 hectopascals at sea level.
Most weather instruments also report sea-level pressure, sometimes labelled SLP or QNH. That is not a direct measurement. It is a calculation: the instrument takes the absolute pressure at its own altitude and adds the pressure that would exist if the same air column extended down to sea level. The result lets stations at different elevations be compared on a weather map. A handheld meter that displays sea-level pressure is doing arithmetic, and the arithmetic depends on the altitude you enter or the altitude the instrument derives from its pressure reading.
That circularity is the reason many handheld weather meters include an altitude correction. The H HOLDPEAK 866B-WM states that altitude is calculated from barometric pressure and that the user can manually correct the barometric pressure before measuring altitude. The listing does not publish the sensor’s absolute accuracy, so the correction improves consistency but does not create a traceable altitude measurement.
Pressure also changes with weather. A deep low can sit near 970 hectopascals, a strong high near 1040. Those are large movements in meteorological terms, but they are small compared with the pressure change caused by climbing a few hundred metres. Near sea level, pressure falls by roughly 12 hectopascals for every 100 metres of elevation, which is why a barometer used as an altimeter needs a reference altitude and why a weather barometer needs to stay at one elevation.
Three sensor types, and the accuracy each can honestly claim
The table below compares the mechanisms on the properties that matter to a buyer. The figures in the first three rows are standard physics or typical sensor behaviour, not claims from the product listings. The product rows use only what the listings publish.
| Property | Mercury column | Aneroid capsule | Silicon sensor |
|---|---|---|---|
| Reference | Weight of mercury column | Elastic deformation of sealed capsule | Piezoresistive or capacitive diaphragm |
| Typical resolution | 0.1 hPa with a vernier | 0.5 to 1 hPa | 0.01 to 0.1 hPa |
| Main drift source | Temperature, mercury purity | Metal fatigue, linkage friction | Temperature, package stress |
| Needs calibration against | Local gravity and temperature correction | A known reference barometer | Factory coefficients, occasionally a reference |
| The BTMETER BT-100WM (BTMETER BT-100WM) | Not applicable | Not applicable | Not stated in the listing bullets |
| The H HOLDPEAK 866B-WM (HoldPeak 866B-WM) | Not applicable | Not applicable | Not stated in the listing bullets; 16-set barometric log stated |
| The Netatmo Smart Weather (Netatmo NWS-AMZ) | Not applicable | Not applicable | Not stated in the listing bullets |
| The ECOWITT GW2001 (Ecowitt GW2001) | Not applicable | Not applicable | Not stated in the listing bullets |
The honest position for all four products is that the listing bullets do not publish a pressure accuracy figure. A buyer who needs a stated accuracy should ask the manufacturer for the calibration certificate or the sensor datasheet. A weather station that reports pressure to one decimal place is not necessarily accurate to one decimal place; resolution and accuracy are different specifications.
Silicon sensors are the most common in consumer instruments because they are inexpensive and small. Their weakness is not resolution but absolute offset, which can be several hectopascals if the sensor is not calibrated. That offset is stable enough for weather trend watching, where you care about change over hours, but it is not good enough for altitude measurement without a local reference.
Where a barometer earns its place, and where it does not
A barometer is the right instrument when you need to observe pressure change over time at a fixed location. Weather watchers use the three-hour tendency, the difference between the current pressure and the pressure three hours ago, to judge whether a system is deepening or filling. A falling tendency below about 1000 hectopascals often precedes unsettled weather in temperate latitudes; a rising tendency suggests improving conditions. The instrument does not predict weather by itself, but it gives you the one variable that moves before the sky does.
Handheld weather meters such as The BTMETER BT-100WM add pressure to wind, temperature, and humidity in one device. That combination is useful for outdoor work where you want a single pocket instrument, but the pressure channel is a secondary function. The listing for the BTMETER BT-100WM states barometric pressure units of inHg and hPa/mbar and a resolution of 0.2 mph for wind, but it does not publish a pressure resolution or accuracy. Treat the pressure reading as indicative, not as a reference.
A barometer is the wrong instrument when you need to detect a gas leak, verify a cleanroom pressure differential, or certify an aircraft altimeter. A gas leak detector responds to a specific gas by catalytic or infrared absorption; a barometer responds to total pressure and cannot tell you which gas is present. A room pressure monitor for a cleanroom or hospital isolation room needs a differential sensor with a resolution of a few pascals and a calibrated zero, not a weather barometer with a resolution of 0.1 hectopascal. An aviation altimeter must be traceable to a national standard and checked against a known altitude; a consumer weather station does not meet that requirement.
For fixed indoor monitoring, a station like The Netatmo Smart Weather reports air pressure alongside temperature, humidity, and air quality. The listing does not publish a pressure accuracy, so the value is in the trend and in the combination with other variables, not in the absolute number. If you need to log pressure over weeks for a building investigation, a dedicated data logger with a published accuracy is a better fit; the principles are the same as those in what an environmental data logger does.
Units, corrections, and the arithmetic that trips people up
Pressure units are a common source of confusion because meteorology, aviation, and engineering each use a different convention. One hectopascal equals one millibar. One inch of mercury equals 33.8639 hectopascals. One pound per square inch equals 68.9476 hectopascals. A reading of 29.92 inHg is the same as 1013.25 hPa. If you compare a station that reports inHg with a forecast that reports hPa, convert before you conclude that the two disagree.
Temperature correction matters for mercury barometers because mercury expands by about 0.018 percent per degree Celsius. A 10 degree change shifts the column height by roughly 0.13 percent, which is about 1.3 hectopascals at standard pressure. Aneroid and silicon sensors have their own temperature coefficients, usually compensated inside the instrument. The listing for The ECOWITT GW2001 states that the outdoor sensor is made from thermal insulation materials and includes a solar panel and backup batteries, but it does not publish a pressure accuracy or a temperature coefficient for the pressure channel.
Altitude correction is the other trap. If you set a handheld meter to sea-level pressure while standing at 500 metres, the instrument will add roughly 60 hectopascals to the absolute reading. That is correct for weather comparison and wrong for altitude. If you then use the same instrument to measure altitude from pressure, you must enter the local sea-level pressure from a reliable source, not the value the instrument just calculated. The circularity is easy to miss.
For logging, the useful specification is not the resolution but the stability over the logging interval. A sensor that drifts by 1 hectopascal over a month will obscure a real weather trend of the same size. The listing for The H HOLDPEAK 866B-WM states that it can record 16 sets of barometric pressure data. That is a manual snapshot log, not a continuous time series, so it is suited to spot checks rather than trend analysis.
Maintenance, drift, and the calibration nobody performs
A mercury barometer needs its reservoir level adjusted, its temperature read, and its reading corrected to standard gravity and standard temperature. That is a laboratory procedure, not a household one. Aneroid barometers need occasional adjustment against a known reference; the small screw on the dial moves the needle, but it does not fix a capsule that has fatigued. Silicon sensors need no user adjustment, but they do need a clean pressure port. A blocked port, a spider in the tube, or a kink in the tubing will slow the response and make the instrument read the pressure inside the enclosure rather than the pressure outside.
Consumer weather stations rarely publish a calibration interval. The honest statement is that the listing does not say. If absolute accuracy matters, compare the instrument against a nearby official station at the same elevation and the same time, and treat any difference as an offset. Do not adjust the instrument unless you have a reference you trust; adjusting to a forecast value introduces the forecast’s error into your instrument.
Battery and power behaviour also affects the record. A station that drops its Wi-Fi connection and reconnects may leave a gap in the log. The ECOWITT GW2001 states a transmission range of up to 150 metres with obstacles and up to 300 metres in open space, which is a useful figure for siting the outdoor sensor. The listing does not publish the logging interval or the memory capacity, so a buyer who needs a continuous record should confirm those before relying on the station.
The maintenance nobody mentions is the comparison itself. A barometer is only as good as the reference you check it against. If you never compare it, you are reading a number, not measuring a pressure. The same discipline applies to humidity and temperature; the reasoning is set out in choosing a temperature and humidity data logger.
The barometer as a reference, not an oracle
A barometer measures the weight of the air column above it. That single number supports weather observation, altitude estimation, and leak testing, but each use needs a different sensor, a different accuracy, and a different reference. The products named here illustrate the range: a handheld weather meter, a logging weather meter, a connected indoor station, and a solar-powered outdoor station. None of them publishes a pressure accuracy in its listing bullets, and none of them is a substitute for a calibrated pressure standard. Choose the instrument for the question you are actually asking, and check it against a reference before you trust the absolute number.
Frequently Asked Questions
Does a barometer measure altitude?
It measures pressure, and altitude is calculated from pressure using a standard atmosphere model. The calculation is only as good as the pressure reference you enter. A handheld meter such as the H HOLDPEAK 866B-WM can correct barometric pressure before measuring altitude, but the listing does not publish an altitude accuracy, so the result is an estimate, not a survey.
Can a barometer predict rain?
It reports pressure and pressure tendency, which are inputs to a forecast, not a forecast. A falling tendency often precedes unsettled weather in temperate latitudes, but the same tendency can occur without rain. Use the barometer with a forecast and with your own observation of the sky.
Why does my barometer disagree with the airport reading?
The airport reports sea-level pressure, which is corrected for elevation. Your instrument may be reporting absolute pressure at your location. At 500 metres the difference is roughly 60 hectopascals. Check which quantity each instrument displays before concluding that either is wrong.
Is a phone barometer accurate enough for weather watching?
A phone sensor is a silicon pressure sensor with a factory calibration, similar in principle to the sensors in consumer weather stations. It is good for trend watching and poor for absolute accuracy unless compared against a reference. The listing for the products here does not publish a pressure accuracy, so the same caution applies.
Can I use a barometer to find a gas leak?
No. A barometer measures total pressure and cannot identify which gas is present. A gas leak detector uses a sensor specific to the gas, such as catalytic combustion or infrared absorption. A pressure drop test can show that a sealed system is losing pressure, but it does not locate the leak or identify the gas.
Do I need to calibrate a consumer weather station?
The listings do not publish a calibration interval. For absolute accuracy, compare the station against a nearby official station at the same elevation and time, and treat any difference as an offset. Do not adjust to a forecast value, because that introduces the forecast's error into your instrument.